TOA-Tube Op-Amp

[bigugly]

I've been thinking about this for a while. What I'd like to do is have a tube opamp that is modular so I can apply it in different ways. ie. pre-amp, virtual earth summing amp, ect.

These are my design goals:

high input Z - pretty easy with tubes
OLG - at least 60db
high CMRR - at least 50db
high PSRR - at least 50db
low THD+N - <0.1%
low output Z - less than 1k

here is what I've got so far.



This is by no means a working design. My problem comes when I think about how to add the buffer stage. The diff. pair is fairly easy and should get around 40db of gain with roughly 1mA of current thru it. Now comes my problem. If I take the inverted output and send it to say a SRPP stage to add more gain and get a lowish output Z then I invariably unbalance the diff stage because I am not loading the two halves the same. I've read that this will hurt my CMRR. I thought about adding a current mirror into the plates of the first stage but I couldn't find any good info for this relating to tubes. (there is alot of info for SS designs) This led me to draw the schem above. My problem with this setup is that I don't know if I have the top tube of the PP pair biased correctly. I did a lot of web surfing and read and re-read "Valve Amplifiers" by Jones but I'm not sure this current configuration will work. Any help would be appreciated.

One more thing. Once I have got this performing to my satisfaction I'm going to post the schem, BOM, PCB layout, and possibly have a run of boards made for myself and fellow forum members. Thanks in advance.

James

 

[gyraf]

..try looking up back issues over at tubecad.com - they had a long article about this once..

Jakob E.

 

[bigugly]

That was quick. Thanks Jakob. I've done a cursory search over there and didn't find much but I'll take a more thurough look again.

 

[gyraf]

Could be me remembering wrong - I can't find it over there either, now.
From what I vaguely remember of the style-of-schematic, it might have been in the magazine "Glass Audio"

Jakob E.

 

[BradAvenson]

http://www.forsselltech.com/schematics/VT%20Opamp.PDF

 

[bigugly]

Thanks Brad, I've looked at both of Fred's designs and learned what I can from them but I am still unsure of how to reach the performance I desire.

 

[CJ]

I have the Tube Tech version if you want it.
And about three million other schemo's that came in from Canada on 14 MP3's thank you very much Mr. B.

I think it would be best not to post the Tube Tech.
I think those guys live near Jake also.

 

[NewYorkDave]

I've said this before and nobody listens, but I'll say it one more time: you can't really make a proper op-amp (for audio use) with only two twin-triodes. You certainly can't meet your target specs that way. Yes, I know the Philbrick K2-W used two 12AX7s, but it's not anything you'd wanna put your audio through. And most of the latter-day "tube op-amp" designs I see floating around on the web are not really proper op-amps.

On the other hand, three twin-triodes would do the job nicely: two triodes for diff pair, one triode for current-source "tail", one triode for common-cathode stage following diff input, two triodes for push-pull output (White CF type).

 

[ciminosound]

[quote author="CJ"]about three million other schemo's that came in from Canada on 14 MP3's thank you very much Mr. B.
[/quote]

Pirates Cove bound I hope!!

:green: :twisted: :green: :twisted: :thumb:

:sam: :sam: :sam: :sam: :sam: :sam:

 

[CJ]

Pirates Cove, what's that?
Oh yeah, never mind.



Arrrhhhh Matey!
OK, on three. Ready?
Everybody:

Yo Ho, Yo Ho! A pirate's life for me.

We pillage, we plunder, we rifle and loot.
Drink up me 'earties, Yo Ho!
We kidnap and ravage and don't give a hoot.
Drink up me 'earties, Yo Ho!

Yo Ho, Yo Ho! A pirate's life for me.

We extort, we pilfer, we filch and sack.
Drink up me 'earties, Yo Ho!
Maraud and embezzle and even hijack.
Drink up me 'earties, Yo Ho!

Yo Ho, Yo Ho! A pirate's life for me.

We kindle and char, inflame and ignite.
Drink up me 'earties, Yo Ho!
We burn up the city, we're really a fright.
Drink up me 'earties, Yo Ho!

Yo Ho, Yo Ho! A pirate's life for me.

We're rascals, scoundrels, villans and knaves.
Drink up me 'earties, Yo Ho!
We're devils and black sheep, really bad eggs!
Drink up me 'earties, Yo Ho!

Yo Ho, Yo Ho! A pirate's life for me.

We're beggars and blighters and ne'er-do-well cads.
Drink up me 'earties, Yo Ho!
Aye! But we're loved by our mommies and dads!
Drink up me 'earties, Yo Ho!

Yo Ho, Yo Ho! A pirate's life for me. :grin:

 

[ciminosound]

Is that sung to the theme of "The Wizard Of Oz"?

:green: :sam: :sam:

 

[bigugly]

First off, thanks for the help eyeryone.

CJ, I'd love to see the TT schem. If you could send it to: biguglyaudio_at_yahoo_dot_com that'd be great.

Jakob, I found the articles over at TubeCad. Very interesting and helpful.

Dave, sorry if I hit a sore spot. I have also come to the conclusion that I'll need at least 3 triodes. Although in a slightly different configuration. The following schematic shows a diff. stage cascaded(?) into another diff. stage which then feeds what is called a "Broskie" cathode follower. (see http://www.tubecad.com/october99/page9.html) Both differential stages have a CCS on their cathodes. The Broskie CF has roughly a 400-500 output Z, which can be decreased further by bypassing its cathode resistors.



So

 

[NewYorkDave]

You didn't hit a sore spot.

The new circuit looks good except I would be concerned about stability when feedback is added. Consider the number of R-C time constants you have in the path--and be sure to include the plate resistance of V1 looking into the Miller capacitance of V2, since that'll be significant enough to warrant consideration

A "Broskie" cathode follower, huh? I'd call it a White with differential drive. Or even a Futterman. Anyway, I'm not familiar with the 5687, but I can tell you that a 12BH7 WCF with about 280V B+ and 15mA of plate current achieves an output Z of about 75 ohms. I've measured it.

 

[BradAvenson]

I would change those NPN's to MPSA42's for a good safety margin. They can see more than 40 Volts before the tube warm up and start conducting.

 

[Samuel Groner]

The new circuit looks good except I would be concerned about stability when feedback is added.

Indeed--it looks to me as if you would have three high-pass filter in series. I guess this makes a good oscillator! Compensation at HF depends on the desired closed-loop gain, obviously.

Samuel

 

[bigugly]

Dave and Sam, yeah it looks like I might run into some problems. Brad, thanks for the tip. I'm going to have to do some more thinking, maybe there is a way to DC couple the stages. Sadly, I have to go and finish putting a roof on my new shop.

btw, Dave thanks for looking over that schem and PCB of the guitar amp. I've changed the biasing to fixed so I was able to eliminate alot of parts and I moved around some resistors so everything seems good now.

 

[alk509]

Yo-ho! 'Tis a fine day to pillage and plunder, me hearties!

We need a goddamned pirate smiley!

Peace,
Al.

 

[ChrioN]

I knew a company who sold TOAs, they used low voltage car stereo tubes if I recall. Can't find the URL though.

 

[ChrioN]

Ok, my bad, it was a hybrid: http://www.welbornelabs.com/hyb.htm

 

[bigugly]

ChrioN, yeah I had loked at that, it is interesting. I haven't progressed much on this. I've been putting a new roof on my shop, but I am investigating DC coupling of the stages.